PHYSICAL REVIEW X 4, 041041 (2014)

Freely Scalable Quantum Technologies Using Cells of 5-to-50 Qubits

with Very Lossy and Noisy Photonic Links
Naomi H. Nickerson,1 Joseph F. Fitzsimons,2 and Simon C. Benjamin3
1

Department of Physics, Imperial College London,

Prince Consort Road, London SW7 2AZ, United Kingdom
2
Singapore University of Technology and Design, 20 Dover Drive, Singapore 138682, Singapore
and Center for Quantum Technologies, National University of Singapore,
Block S15, 3 Science Drive 2, Singapore 117543, Singapore
3
Department of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, United Kingdom
(Received 25 July 2014; published 9 December 2014)
Exquisite quantum control has now been achieved in small ion traps, in nitrogen-vacancy centers and in
superconducting qubit clusters. We can regard such a system as a universal cell with diverse technological
uses from communication to large-scale computing, provided that the cell is able to network with others
and overcome any noise in the interlinks. Here, we show that loss-tolerant entanglement purification
makes quantum computing feasible with the noisy and lossy links that are realistic today: With a modestly
complex cell design, and using a surface code protocol with a network noise threshold of 13.3%, we find
that interlinks that attempt entanglement at a rate of 2 MHz but suffer 98% photon loss can result in
kilohertz computer clock speeds (i.e., rate of high-fidelity stabilizer measurements). Improved links would
dramatically increase the clock speed. Our simulations employ local gates of a fidelity already achieved in
ion trap devices.
DOI: 10.1103/PhysRevX.4.041041

Subject Areas: Quantum Physics

Within the past year, there have been remarkable

advances in the fidelity with which small quantum devices
can be controlled. The two most mature systems are ion
traps and superconducting qubits. In ion trap devices,
single-qubit fidelities [1] have reached 99.9999%, with
combined preparation and measurement of 99.93%.
Moreover, two-qubit operations [2] have been reported
with fidelities up to 99.9%. Meanwhile, a superconducting
qubit device (SQD) containing five qubits [3] has been
demonstrated with all qubit manipulations above 99.3%.
At the same time, there has been rapid progress in the study
of nitrogen-vacancy (NV) centers in diamondsingle
electron spin manipulation is possible with 99% fidelity
[4], and it is possible to manipulate nuclei that are relatively
far from the center, so that each NV center may be thought
of as a group of several qubits interacting with an optically
active core [5].
These prototype systems are small; none of them contain
as many as 20 qubits. But importantly in each case, it is
possible to bridge between small systems using photonic
channels, albeit with lower entanglement fidelities and in a
probabilistic way that may require many attempts. In the
ion trap community, there are well-established methods for

Published by the American Physical Society under the terms of

the Creative Commons Attribution 3.0 License. Further distribution of this work must maintain attribution to the author(s) and
the published articles title, journal citation, and DOI.

2160-3308=14=4(4)=041041(17)

entangling ions in separate traps, and recent progress

associated with projects such as the MUSIQC initiative
[6] have led to successful entanglement at a rate of hertz [7].
This can be improved by orders of magnitude by hardware
advances and by loss-adapted protocols, as we describe in
this paper. In SQDs, a well-established means of interfacing
qubits is to exploit microwave photons in cavities [8]. This
suffices for short-range bridging, and, moreover, remote
entanglement of two superconducting qubits separated by
more than a meter of coaxial cable has recently been
demonstrated [9]. In the case of NV center research,
successful optical linking of qubits can occur either within
the same sample [4] or over meters of separation [10],
and teleportation [11] with fidelity around 86% as been
achieved.
Thus, the quantum state of the art includes wellcontrolled small groups of qubits, which we refer to as
cells in this paper, together with intercell entanglement
links that may be nondeterministic and, even when successful, noisy. These ingredients may already suffice to
develop fully scalable technologies: although the fidelities
over the links are too low to directly enable secure
communication or fault-tolerant computing, crucially the
fidelities within cells are now high enough to support
entanglement purification. This process allows one to
improve the fidelity of a quantum channel by combining
several successive uses of the link. Thus, at the cost of
lowering the effective bandwidth, we have a powerful
paradigm in which small cells link to one another through a

041041-1

Published by the American Physical Society

NICKERSON, FITZSIMONS, AND BENJAMIN

PHYS. REV. X 4, 041041 (2014)

FIG. 1. A small, well-controlled quantum system interfaced to a

noisy entanglement-sharing channel constitutes a universal cell if
it can purify the entanglement to a high fidelity. Such cells enable
secure communication; monogamy of entanglement [12] means
the links need not be secure and may be either direct (i) or via
repeater hubs (ii) or switches (iii). (b) Moreover, a dense array of
cells bridged by short links constitutes a freely scalable computer,
as we analyze here.

kind of internal digital filter where purification is performed; see Fig. 1. This paradigm universally supports
quantum technologies on any scale. On the large scale,
when the bridges between cells are meters or kilometers
long, cellular nodes enable secure communication and
other distributed information tasks. However, this paper
reports on freely scalable quantum computing, where the
optical bridges connect a dense array of cells with spacings
on the order of centimeters or less.
Any technology based on high-performing cells bridged
by very imperfect links will be practical only if entanglement purification is efficient and robust. The protocol
should have frugal requirements for work space qubits
within a cell, it should require only achievable levels of
fidelity for local gates and measurements, and, most
importantly, it should minimize the time cost by requiring
only a few uses of the noisy quantum channel (i.e., a small
number of low-fidelity Bell pairs) in order to purify a highfidelity shared state. These desiderata are in tension with
one another, and, of course, the achievable values will
depend on the fidelity of the native channel as well as the
target fidelity that enables the task in question (e.g.,
communication or computation). Furthermore, certain
tasks, such as the stabilizer-based computing considered
here, are best enabled by multiparty entangled states that
are more complex than simple Bell pairs.
A seminal Letter in the purification literature is that of
Briegel and Dr [13], which showed that by using a tiering

system, one can promote even very noisy raw entanglement to a fidelity that is of the order of the fidelity of the
local operations. A number of authors have extended this
idea, for example, through a means to purify phase noise
very efficiently [14] and by introducing the idea of double
selection [15], which was then used in a scheme [16] that
can tolerate channel noise up to 30% in the context of
quantum computing. Recent work has even pushed the
acceptable limits of local noise to comparable levels [17].
Here, we employ a range of such techniques in order to
obtain what we believe to be the most practical purification
protocols yet described for the context of quantum computing. Moreover, we show how these protocols can guide
specific hardware design to achieve optimum efficiency.
We aim to determine whether the intracell operations and
limited intercell links that are regarded as achievable today
can suffice for full-scale quantum computing, assuming
that the various accomplishments that have been made in
different (but compatible) experiments can be engineered
into a single platform. We conclude that the answer is yes.
Our approach here is an evolution of the scheme in
Ref. [18], which was designed to fight network noise. We
extend that scheme to be efficient versus severe loss while
retaining the noise tolerance, and this allows us to analyze
the clock speed of the resulting computer. Our approach
requires a total of at least five qubits per cell: four qubits
for purification of noise on the cell-cell coupling links and
one that is involved in the actual quantum computation
(a so-called data qubit, as we explain presently). Ion
traps, SQDs, and NV center systems can all scale to five
qubits. However, with ion traps and SQDs, we may
eventually have the luxury of tens of qubits per cell. In
that case, we can make good use of the additional structure:
presently, we discuss a buffered ion trap design that is
optimal for entanglement purification, maximizing the
processing speed of a computer formed from such cells.
A cell possessing tens of qubits could also embody multiple
data qubitswe do not pursue this possibility here, but
it is an obvious method for reducing the number of cells
required for a given computational task. In any case, a
useful quantum computer will require a great many basic
cells, but since each cell is likely to be of subcentimeter
dimensions, a machine composed of millions of cells
could fit within the space allocated to a conventional
supercomputer.
To support quantum computing with the cellular paradigm, we must select an approach to achieving fault
tolerancethis will effectively set the target fidelity with
which purified intercell operations must be performed. We
opt to employ a surface code, first introduced by Kitaev
and co-workers [19,20], because of its high thresholds
and local structure [21]. The approach involves repeatedly
measuring certain stabilizersthese correspond to simple
parity measurements on groups of qubits; i.e., we need to
learn whether the total number of 1s in the group is odd or

041041-2

FREELY SCALABLE QUANTUM TECHNOLOGIES USING

PHYS. REV. X 4, 041041 (2014)

FIG. 2. Monolithic versus network quantum computing. A piece of surface code is represented in the monolithic picture on the left,
and in a network architecture on the right. (Here, for clarity we depict a 2D planar code; our network simulations will actually employ
the toric code, i.e., periodic boundaries. These two variants are very similar; see Appendix B.) In a monolithic structure [24,25],
all qubits are contained within one physical system and two-qubit gates are performed directly between qubits via some physical
interaction. In the network picture, the system is instead divided up into small cells, each of which contains a modest number of qubits
that interact directly, while between cells only noisy and lossy interactions are possible. For the monolithic system, stabilizer
measurements are performed by using a dedicated set of ancilla qubits (pink) interlaced with the data qubits (blue). In contrast, an
efficient route to making stabilizer measurements in the network model is to first purify a shared GHZ state between the cells involved
and then use this resource to evaluate the stabilizer: The parity of the four qubits measured out from the GHZ tells us the stabilizer
outcome. In subsequent cycles, cells are grouped into different sets of four in order to evaluate a complementary set of stabilizers; see
Appendix B 3.

even. The basic repeating cycle of the computer involves

alternating patterns of parity checks separated by Hadamard
rotations to switch between the x and z basis. Remarkably,
this simple principle allows for far more than merely
protecting quantum information from errors: certain operations between encoded logical qubits can be performed
merely by altering the patterns of parity measurements
[20,22], and together with a technique such as magic state
distillation [23], all the operations required for universal
quantum computation can be performed this way.
In a monolithic 2D device, there is a natural layout for
the physical qubits such that nearest-neighbor interactions
suffice to efficiently perform the parity evaluation (see
Fig. 2, left). But how should one find the parity of four data
qubits if they are instead incorporated into four different
cells? The solution used in Ref. [18] is to employ ancilla
qubits within the cells, interacting them with one another
across the network so as to build up a four-qubit
Greenberger-Horne-Zeilinger (GHZ) state with one qubit
in each cell (see Fig. 2, right). Given this GHZ state, it is
trivial to deterministically find the parity of the data qubits,
as shown in the figure. The challenge is to efficiently
make a high-fidelity GHZ state in an efficient manner. We
now specify the protocols that we have developed (Fig. 3)
and we establish their performance in terms of the faulttolerance thresholds. The full process of deriving these
performance figures is fairly involved, and is described in
Appendix C.
In order to minimize the impact of photon loss on
entanglement generation rates (and so ultimately maximize
the computers clock speed), we must optimize the mechanism by which raw entanglement is achieved between cells.
Typical schemes for optical entanglement generation, such

FIG. 3. Distilling a GHZ state with EPL-generated entanglement. (a) The symbols with gray-shaded stars represent the highly
mixed Bell states raw or 0raw obtained when a single detector
click is seen. Following the the extreme photon loss protocol
[26], we can combine two such pairs and measure one out,
thus producing a single greatly improved pair. This process is
represented by a symbol with an open star. (b) Circuit diagram
showing the adapted GHZ distillation process for entanglement
generated using the EPL protocol. Two GHZ states are produced
and one is used to make a four-qubit parity projection onto the
other. The three different protocols, basic, medium, and refined
are shown.

041041-3

NICKERSON, FITZSIMONS, AND BENJAMIN

PHYS. REV. X 4, 041041 (2014)

as the Barrett-Kok method [27], are based on heralding by

two photons so that the rate of successful entanglement has
a quadratic dependence on the probability that a photon
avoids loss, RBK 12 1 ploss 2 . Here, ploss is the probability that an emitted photon fails to make it through the
system and yield a detector click, whether due to loss or to
detector failure. The antibunching scheme employed by
Monroe and co-workers in Ref. [7] has a similar form:
RM 14 1 ploss 2 . In both cases, the quadratic dependence is punishing when losses are severe and only rare
photons are captured and detected. The use of cavities to
enhance matter-light coupling may eventually allow more
sophisticated entanglement channels (as recently demonstrated [28,29]), but here we assume that cavities are not
employed, and therefore, we must minimize the impact of
loss. We adapt a scheme of Campbell and Benjamin [26]
called the extreme photon loss (EPL) protocol, which
requires one additional qubit at each site and results in
success rate REPL 18 1 ploss , i.e., linear in the photon
loss rate (the precise prefactor depends on a parameter in
the scheme).
In common with the approach of Barrett and Kok, this
scheme requires a system capable of conditionally emitting
a photon depending on its state (see lower left-hand panel in
Fig. 1). Two optically active broker qubits [30], each in a
p
p
separate cell, are initialized to a state p0 j0i p1 j1i and
optically excited causing the j1i component to emit a
photon as it immediately decays from a short-lived excited
state. Any such photons emitted pass through a beam
splitter before impinging on photon detectors. The EPL
protocol assumes that useful entanglement is heralded by
the detection of a single photon click. In the absence
of photon loss or noise, this would p
produce
an odd parity

Bell state, say, ji j01i j10i= 2. In reality, there are

various sources of error in this process. The first is the
consequence of general imperfections in the preparation
and manipulation of the qubits, which we model by mixing
the ideal Bell state with the identity, leading to
imperfect 1 pn jihj

pn X
j ih j;
3 i1;2;3 i i

where the i are the other three Bell states. We note that
the noise model is actually quite general: any state can be
twirled into this form using local operations, which (being
relatively high fidelity) will not significantly degrade the
entanglement. However, if the imperfections in the system
lead to a state with biased noise, as, for example, if phase
noise dominates, then in fact this bias may be advantageous
[14]; it is typically most difficult to purify structureless
white noise of the form assumed here.
Now, in addition to this general noise, we have the
specific problems of photon loss and dark counts. With
photon loss, the primary issue is that when we see a single
click, as required by the protocol, it may be that that, in fact,

two photons were emitted, one from each qubit, but one
photon was lost, and we thus incorrectly heralded a success.
In that event, the eventual broker state is j11i. Because of
this possibly, our state is
raw 1 rimperfect rj11ih11j;

where r ploss =p1

1 1 ploss . In our simulations, we
assume that ploss is very severeit approaches unity and,
therefore, r p1 . Thus, raw is highly mixed: its two terms
will have comparable weight.
Having thus accounted for photon loss, it remains to
assess the impact of dark counts. Our protocol proves to
be quite robust versus this issue; a full analysis is presented
in Appendix F. We show that the key parameter is
d pdc =1 ploss , where pdc is the probability that a
given detector registers a dark count in the detection
window of a single entanglement attempt. Provided that
d 102 , then to a good approximation dark counts simply
increase the network infidelity; for example, if we set
r 1=2, then finite dark counts result in pn pn 3d.
Given that we have seen a detector click and so heralded
the existence of raw , we now store this state and proceed to
create another instance of it. Note that there will typically
be many heralded failures, i.e., instances where no detector
click is reported, before another success is seen. When that
success occurs, we again have a raw , except that we apply
an additional random phase shift to account for the fact that
a substantial time may pass between creation of the two
pairs, so that a finite unknown phase drift in the network
may have occurred. This case be can modeled by saying
that a phase shift has occurred with probability pdrift, as
follows:
0raw 1 pdrift raw pdrift Z1 raw Z1 :

Note that apart from this possible drift between the two
heralded successes, the approach is otherwise interferometrically stable: an unknown phase shift that is acquired
by both raw and 0raw will cancel out in the next step; see
Appendix A. This step proceeds as shown in Fig. 3. Within
each of the two cells, a local control-NOT operation is
performed; it is controlled by the broker associated with
raw and targets the broker associated with 0raw . The
brokers associated with 0raw are then separately measured
in the z basis. As explained in Appendix A, the measurement outcome 1,1 is inconsistent with either of the two
pairs raw or 0raw having been originally in the j11i state;
thus, if that outcome is seen, the j11ih11j component of the
surviving entangled pair is removed. A convenient feature
of this protocol is that the desired measurement outcome,
j11i, can be made to correspond to bright states of the
matter qubits, which have a higher measurement fidelity
than their dark counterpart in several optical systems.
Of course, the local intracell operations and measurements must themselves be treated as noisy (see Appendix C

041041-4

FREELY SCALABLE QUANTUM TECHNOLOGIES USING

for the noise model). Given that they are reasonably high
fidelity, the result of a successful 1,1 outcome is the Bell
state EPL which, while still imperfect, is far higher fidelity
than the parent states raw and 0raw . We then take these
EPL-derived Bell pairs EPL as the basic resource for our
GHZ creation. (Optionally, we could use the EPL protocol
to perform some, or all, of the parity projections involved in
creating the GHZ; however, this possibility is not explored
here.) We introduce three new purification protocols with
varying time-versus-fidelity tradeoffs. These are depicted in
Fig. 3(b). The basic protocol is fast but tolerates only a
limited network error rate. The refined protocol carries out
several rounds of entanglement distillation, making it much
more robust against noise, but also quite time consuming.
The medium option sits between these two extremes.
Having established our procedure for generating shared
GHZ states across the network links, we proceed to
determine the performance of the quantum machine by
simulating and tracking errors. This is an intensive numerical
process benefiting from the use of a cluster-scale computer
facility. The process is detailed in Appendix C. We summarize it here: For a given set of local error rates, we pick a
network error rate pn and a network size characterized by
parameter L such that there are 2L2 cells in the complete
toric network. We then simulate a large number of stabilizer
cycles of the computer. At the end of this simulation, we
inspect the state to determine whether the logical qubit
was corrupted, a simple yes or no outcome. We repeat this
numerical experiment many thousands of times (typically
3 104 ) to determine the probability that logically encoded
qubits will survive these stabilizer cycles without error;
this produces one data point for Fig. 4. This process is now
repeated with a different network sizeif the larger network
has a lower logical qubit error rate, we deem that the surface
code is operating successfully and, therefore, our chosen
network error rate is within the threshold for fault tolerance.

PHYS. REV. X 4, 041041 (2014)

The analysis is then repeated for different levels of network

noise in order to determine the threshold precisely.
The results of these calculations are shown in Fig. 4,
from which we find that the basic protocol leads to a
threshold of 7.7%, the medium complexity protocol has a
threshold of 13.3%, while the most aggressive protocol,
refined, is able to tolerate very high network noise of up to
19.4%; note in each case we allow for an additional 1%
phase drift between the two rounds of EPL.
The question of how these protocols behave when well
below threshold (i.e., the regime where a device would
realistically operate) requires a different approach to the
Monte Carlo simulations performed here, as considered
in several recent works [31,32]. While this is beyond the
scope of this paper, we note that using the medium protocol
at half the threshold network error (7%) with a lattice size
of L 16 yields a logical error rate per L stabilizer rounds
of fewer than 1 in 106 .
These simulations establish the tolerable levels of error,
which are comparable to (but better than) our earlier paper,
Ref. [18]. However, because this new approach is founded
on the EPL protocol for entanglement generation, the
overall time needed to perform a stabilizerand, hence,
the fundamental clock cycle of the quantum computer
will be much faster than in prior schemes.
The achievable computer speed depends on the kind of
cell architecture that we have available; see Figs. 5 and 6.
An obvious advantage is to have multiple entanglement
channels connected to each cell; the example in Fig. 5(c)
has eight channels (two to each of its four neighboring
cells; see also Fig. 11). There is also a second, independent
characteristic of the architecture, which we analyze by
distinguishing two limiting cases, the minimal architecture
and the buffered architecture. In a minimal system, there are
just enough qubits to perform our protocols. There will then
be uncertainty as to how long it takes to complete a given

FIG. 4. Results of threshold calculations for the three protocols considered: (a) basic, (b) medium, and (c) refined. The logical error
rate in the toric code is calculated for varying values of the network error rate pn . The cells internal error rates are taken to be the best
currently demonstrated in an ion trap system: a two-qubit gate error rate of 0.1% and a measurement error rate of 0.05% [1,2]. We select
p1 14 r 14, and we take the phase shift in the EPL entanglement generation to be pdrift , of 1%. Details of the error model can be
found in Appendix C. The three curves on each plot denote the results for increasing lattice sizes, where L 8, 12, and 16 (therefore
containing 2L2 data qubits). The threshold is defined as the intersection of these curves from which we find the basic protocol has a
threshold of pn 7.7%, medium has a threshold of pn 13.3%, and refined has a threshold of pn 19.4%. In Appendix E, we further
explore the dependence of the network error threshold on the error rates for intracell two-qubit gates and measurements.

041041-5

NICKERSON, FITZSIMONS, AND BENJAMIN

PHYS. REV. X 4, 041041 (2014)

TABLE I. The threshold of tolerable network error rates for
each of the three distillation protocols considered, and the time
cost for making a complete high-fidelity four-qubit GHZ assuming we operate well under threshold (3%, 5%, and 7% for the
three protocols, respectively). Such a GHZ state enables a
stabilizer measurement. The distinction between minimal and
buffered architectures is illustrated in Fig. 5.
Time to make GHZ
(units of T 0 )
Protocol

Threshold error rate

(pn pdrift )

Minimal
architecture

Buffered
architecture

Basic
Medium
Refined

7.7% 1%
13.3% 1%
19.4% 1%

22
47
102

5.2
12.2
31.6

cost to perform a high-fidelity stabilizer measurement on

four data qubits. It is quantified in terms of T 0 , the time to
produce a single basic EPL Bell pair, i.e., the state EPL .
Given all of these contributing factors, we can now
estimate an achievable rate for the clock cycle of our
computer. We neglect the time for local gates and measurements; thus, our estimate is accurate only if such gates

FIG. 5. Example architectures relevant to the cellular network

paradigm. In (a) and (b) there are 5 qubits available, the minimum
required by the protocol: (a) An NV center with several 13 C atoms
within range of the core constitutes a five-qubit cell with one
qubit (the electron spin) coupling to an optical channel. A simple
ion trap (b) need only have five ions, but a more complex
architecture (c) offers the advantage of temporarily storing, or
buffering, the incomplete GHZ states. In this illustration, the
eight independent entanglement sites further enhance the GHZ
generation rate and thus increase the clock speed of the
computer; this also obviates the need for optical switching, as
illustrated in Fig. 6. The small square symbols indicate the
generation of Bell pairs between cells and the subsequent
synthesis of GHZ states out of those Bell pairs. A filled circle
indicates an ion in this cell, whereas open circles are ions in
neighboring traps; this may be more apparent from the multicell
schematic in Fig. 11.

stabilizer measurement (since the protocols are probabilistic) and this necessitates a delay for synchronization; see
Appendix D 1. In contrast, a buffered architecture has
additional internal storage allowing us to queue our qubits
and smooth out the timing irregularities, thus avoiding the
difficulty in synchronization. Table I summarizes the time

FIG. 6. Illustration of the cellular architecture in 3D. See also

the detailed schematics in Figs. 10 and 11.

041041-6

FREELY SCALABLE QUANTUM TECHNOLOGIES USING

are performed on the scale of microseconds. This appears
achievable, but we note that in established experiments, the
highest fidelities are seen for longer gate times (see, e.g.,
Ref. [2]). One cycle of a surface code quantum computer
corresponds to a set of parity measurements over all its data
qubits (either in the x basis or the z basis, alternatingly).
Suppose that the cells in our machine correspond to the
design in Fig. 5(c), or a superconducting qubit device of
equivalent complexity. Assume that each cell-cell link is
an entanglement channel, which is realistic with todays
technology: the entanglement attempt rate is 2 MHz and the
end-to-end photon detection probability is only 2%. We
select p1 14 and find that the average time cost for an
entanglement channel to create one Bell resource (EPL ) is
T 0 0.27 ms. Now, further assume that we have opted for
the medium purification protocol because we have network
noise at a level of 5% (well within the medium protocols
threshold of 13.3%). According to Table I, a single-channel
cell will require time 12.1T 0 to create one high-fidelity
GHZ state. Our cells have eight channels, which together
generate such GHZs at a rate of 2.5 kHz; however, two
GHZ states per cell are consumed in making a complete
set of stabilizer measurements (either x basis or z basis), as
explained in Appendix C 3. Therefore, our overall clock
rate is 1.2 kHz.
Higher rates could be achieved simply by introducing
more entanglement channels [the branched design in
Fig. 5(c) obviously generalizes from 8 channels to 2N ].
This is consistent with ideas in the MUSIQC project [6].
Alternatively, if we look to the medium-term future and
assume that the use of integrated cavities [29] (or other
advances) can reduce the photon loss rate to 50%, and that
the network noise can be taken well below the 7.5%
threshold of our basic protocol, then the same device
design in Fig. 5(c) should begin to approach megahertz
rates for stabilizer measurement. At this point, the local
gate speeds may be the limiting factor.
In conclusion, we consider an architecture for quantum
technologies that is motivated by the recent achievements
in ion traps, superconducting devices, and NV centers.
We consider small quantum cells composed of a few
(5-to-50) qubits under high-fidelity deterministic control,
together with intercell network links that are both noisy
and lossy and therefore nondeterministic. This architecture
is relevant to communication when the links are long (e.g.,
kilometers), but our focus here is on quantum computing
using a large number of such cells with short (e.g.,
centimeter) bridges. We find that the exquisitely high
levels of control recently achieved in small systems can
enable very compact and efficient entanglement purification, allowing one to use relatively poor photonic links
between cells. We consider three different purification
protocols and derive the corresponding thresholds for
fault-tolerant quantum computing, finding that threshold
network fidelity can be as low as 80%. Moreover, we study

PHYS. REV. X 4, 041041 (2014)

the time cost of the purification process, and thus the time
to evaluate a set of stabilizer measurements across the
networkeffectively, the clock speed of the quantum
computer. We relate this speed to the complexity of the cell.
Given cells that are sufficiently complex to incorporate
parallel operations and buffering (temporary storage),
we find that even the highly lossy links that are realistic
today should support kilohertz-rate, freely scalable quantum computing.
ACKNOWLEDGMENTS
We acknowledge very useful conversations with Chris
Ballance, Earl Campbell, Austin Fowler, Ronald Hanson,
Tom Harty, Winfried Hensinger, and David Lucas.
Computing facilities were provided by the Advanced
Research Computing service of the Univeristy of
Oxford, the Imperial College High Performance
Computing Service, and the Dieter Jaksch group. S. C.
B. acknowledges EPSRC platform Grant No. EP/J015067/
1. This material is based on research funded in part by the
Singapore National Research Foundation under NRF
Award No. NRF-NRFF2013-01.
APPENDIX A: EXPLANATION
OF THE EPL METHOD
Here, we describe the EPL protocols handling of photon
loss and its inherent tolerance of systematic phase errors.
We neglect all other imperfections, but of course these
are accounted for elsewhere in our analysis and in our
simulations (cf. Appendixes C 1 and F).
For simplicity, we initialize each of our
p two remote
broker qubits into the state j0i j1i= 2. (Note that,
in fact, we need not start from an equal superposition;
one selects an optimal level of excitation.) The j1i state is
optically active; we excite both brokers and route the
collected light through a beam splitter prior to detection
as in the standard picture. If we see a photon, then we have
simple 1 rjihj rj11ih11j;
p
where ji j01i ei j10i= 2 and is some phase
introduced due to the optical apparatus. The ratio r depends
on the severity of photon loss; in the limit of loss tending to
unity, we find r 12. An intuitive explanation of this is as
follows: the state ji generates only one photon whereas
j11i generates two; however, the state ji has twice the
probability of j11i in the original broker-broker product
state, thus these factors compensate and the two states have
equal weight in the mixture. Now we store simple and
attempt to create another such pair. We may fail a number
of times before we again succeed. Provided that the
apparatus does not suffer phase drift between the two
successful events, the second entangled pair will have the
exact same form, simple . (Recall that our full analysis in the

041041-7

NICKERSON, FITZSIMONS, AND BENJAMIN

PHYS. REV. X 4, 041041 (2014)

main text does account for the possibility that such a drift
occurs.)
Having obtained two noisy entangled pairs, each of the
form simple , we now apply CNOT gates locally within each
cell according to the Fig. 3(a) circuit. These CNOT gates
map each of the pure states within our simple simple
mixture as follows, where we take the left-hand side to be
the controlling qubits and the right-hand side as the target
qubits:
j01i ei j10ij01i ei j10i
j01ij00i ei j11i ei j10ij11i ei j00i
j11ij01i ei j10i j11ij10i ei j01i

fact, the value of a threshold obtained from the two

approaches (given all other factors are held the same)
varies only slightly.
APPENDIX C: SIMULATION METHODS
Here, we give an overview of the methods used in
calculating threshold values for our system. The approach
can be divided into two distinct sections: In the first part,
we derive superoperators representing the net effect on the
data qubits of our stabilizer measurement protocols with all
their various errors. In the second part, we use a classical
algorithm to track the effect of these superoperators as we
simulate a surface code embodying logical qubits.

j01i ei j10ij11i j01ij10i ei j10ij01i

j11ij11i j11ij00i:

1. Error model

A1

The protocol then calls for the one of the qubit pairs, the
right-hand side pair in the present notation, to be measured
in the z basis. Any result other than 11 is rejected. We see
that only the first of the four possibilities listed above
can pass this filter. Revisiting Eq. (A1) and collecting
terms, we have
j01ij00i ei j11i ei j10ij11i ei j00i
j01i e2i j10ij00i ei j01i j10ij11i:
Thus, measuring 11 implies
that the remaining qubit pair
p
is state j01i j10i= 2 and we have eliminated both
the j11i components due to photon loss and the unwanted
phase (which, therefore, we need not know). This occurs
with probability 12 1 r2, i.e., 18 when r 12.
In practice, of course, there are other sources of error,
both in the network and in the local gates and measurements, as discussed in the error model below. But for all
levels of noise relevant to the devices we are considering,
the result of this process is to generate a state EPL , which is
far higher fidelity than the two parent states.
APPENDIX B: PLANAR VERSUS TORIC
NETWORK TOPOLOGIES
The main text presents results for the toric code. In this
version of the surface code the boundaries of the surface
wrap to form a torus. While this would be difficult to realize
if we were employing a monolithic structure (cf. Fig. 2,
left), in a network paradigm there is no in-principle
difficulty. However, it might be that there are reasons to
prefer to layout the network in 2D and maintain all links of
the same physical lengthin this case, one would adopt the
planar version of the surface code. Our threshold-finding
numerical programs can simulate either the toric or the
planar variant. We find that while the toric code exhibits a
slightly sharper threshold because it has no edge effect, in

All of the protocols we consider are composed of a small

number of low-level basic operations, each with an associated noise model. 1. Network error model. This is
described in the main text. States raw and 0raw are defined
in Eqs. (2) and (3). They involve the network noise rate pn
and the photon loss rate ploss , as well as a parameter pdrift
accounting for phase drift in the entanglement channel
between the creation of the two states. The dark-count
rate d can be subsumed into the network noise pn as we
explain in Appendix F. 2. Local (intracell) controlled-Z and
controlled-X gates. For a gate error rate pg , the noise is
modeled as a perfect gate operation, which with probability
pg is followed at random with one of the 15 nontrivial
two-qubit Pauli errors i j , where i 0; 1; 2; 3 and 0
is the identity. If represents the ideal state after gate
operation, this noise map can be written as
N gate 1 pg

pg X
j i j :
15 i;j0;0 i
C1

3. Single-qubit measurement in the X and Z bases. Given

a measurement error rate pm , a particular outcome of
the measurement, q f0; 1g, corresponds to the intended
projection Pq applied to the state with probability (1 pm )
and the opposite projection Pq applied with probability pm.
This noisy projector can be written as
qj:

P q pm 1 pm jqihqj pm jqih

C2

2. Stabilizers as superoperators
To characterise the entire process of the stabilizer
measurement, we carry out a full simulation of the
measurement procedure including all sources of noise
and use the Choi-Jamiolkowski isomorphism [33] to
generate a superoperator from the result. Thus, we completely describe the action of the stabilizer measurement
procedure with

041041-8

FREELY SCALABLE QUANTUM TECHNOLOGIES USING

S

X
pi K i K i :

C3

i0

This probabilistic decomposition describes the operation

as a series of Kraus operators K i applied to the initial state
with probabilities pi , which depend on the chosen protocol,
noise model, and the error rates. The leading term i 0
will have corresponding K 0 representing the reported parity
projection, and large p0 . For the protocols considered
here, the other Kraus operations can be decomposed and
expressed as a parity projection with additional erroneous
operations applied. For example, if a noisy stabilizer
measurement is made that returns an even outcome,
we find K 0 Peven , the reported even parity projection,
and K 1 Podd , which implies that a perfect odd parity
projection was applied, but the wrong outcome was
recordeda lying stabilizer measurement. All of the
other K i can be represented as K 0 or K 1 followed by singlequbit Pauli errors. This decomposition then involves two
distinct types of error: lies, where an incorrect outcome is
recorded, and qubit errors, where a physical error occurs on
a data qubit. The probability of each combination of events
can be calculated from the values of the pi . This information on stabilizer performance then enables classical
simulation of a full planar code array, and its fault-tolerance
threshold can be assessed.
3. Scheduling stabilizer measurement
Each qubit in the body of the lattice is part of four
different stabilizer groups. Therefore, in a physical implementation, the measurement of a full cycle of stabilizers is
divided into four distinct rounds, two of plaquette measurement and two of star measurement, as shown in Fig. 7.

PHYS. REV. X 4, 041041 (2014)

For the purpose of simulation, it is desirable to break down

the evolution of the lattice into complete rounds of perfect
plaquette or star measurement separated by rounds of
errors. This can be achieved by making use of the fact
that each Kraus operator can be decomposed in different
but equivalent ways, namely, that each K i can be written
with the Pauli errors either preceding or following a parity
projection.
4. Decoding
Decoding is performed using Kolgomorovs Blossom V
implementation of Edmonds minimum weight-matching
algorithm [34,35] to generate a perfect matching
between stabilizer violations. To do this, the syndrome
on the lattice must be formulated into a weighted graph.
In the case of perfect measurements, each 1 outcome
in the syndrome becomes a node of a completely
connected graph, where the weight of each edge is given
by the distance between the corresponding two nodes on
the lattice.
Multiple rounds of stabilization are performed, producing a three-dimensional syndrome cube where the third
dimension represents time. Each point where one stabilizer
measurement differs from its value in the previous round
gives rise to a node in the graph. Matching in the spatial
dimensions of the cube correct for physical errors on the
lattice, while timelike matchings correct for lying stabilizers. The rate of lie-type errors and physical errors will not
generally be the same; to account for this, timelike and
spacelike paths are weighted differently. The ratio of these
weights is chosen to optimize performance.
In the toric code, error chains on the surface will always
result in two stabilizer violations, one at each end of the

FIG. 7. Scheduling stabilizer measurements. Stabilizer measurements are split into four rounds, two of each type, plaquettes (orange)
and stars (blue). The underlying lattice and qubit structure is shown here in black for a planar code of lattice dimension L 4.
Decomposition of the stabilizer measurement procedure allows each round to be described as a round of perfect measurements either
followed or preceded by errors. The stabilizer implementation including these errors is shown on the right.

041041-9

NICKERSON, FITZSIMONS, AND BENJAMIN

PHYS. REV. X 4, 041041 (2014)

chain. In the planar code, however, if an error chain reaches

an edge of the lattice, only one stabilizer violation will be
seen. To account for this, each node in the original graph is
uniquely connected to a new node located at the nearest
boundary position, following the method described in
Ref. [36]. This gives the possibility for each 1 stabilizer
to match to a boundary as well as any other node on the
lattice itself.
The process described above is a vanilla implementation of the perfect-matching decoderthere are many
possibilities for optimizing the decoder that are not
pursued in this paper. For example, it is well understood
that correlating the X and Z errors reveals information
about Y errors. Moreover, there are opportunities to
exploit the classical information that occurs during a
stabilizer evaluation: Most importantly, in the case where
we are using a simple, serial architecture with no buffering
available (a NV-center-based technology, for example),
we need to impose a cutoff time after which an attempt
to measure a stabilizer is abandoned; see Appendix D 1.
At present, our decoder makes no use of the information that a given stabilizer has not been evaluated,
and instead simply replaces the missing information with
a copy of the previous resultthis is obviously not
optimal.
We emphasize that these limitations in our decoder do
not undermine the accuracy of the simulations in this paper,
since the operator of a quantum computer is free to use any
classical decoder they wish when they run the machine.
Thus, the thresholds that we find should be considered
a lower bound on the achievable thresholdsa better
decoder should boost the performance.
5. Calculating the threshold
We wish to determine the logical error rate of the code
as a function of the physical error rates. To do this, we
perform Monte Carlo simulations of a lattice under noisy
stabilizer measurement. For each instance, the evolution
of the lattice is simulated by applying random errors
drawn from the distributions specified by the derived
superoperators. A total of 3L complete stabilizer rounds
are performed, where L is the lattice dimension, before a
decoding attempt is made and the result analyzed to test
whether a logical error has occurred. For each physical
error rate, the logical error rate is calculated for three
different lattice sizes, L 8, 12, 16. For each data point,
a minimum of 30 000 instances are simulated, and error
bars are calculated by treating each result as a sample
drawn from a Bernoulli distribution.
If the error rate is below the threshold, then increasing
the lattice size will improve the performance of the code;
that is, the logical error rate will become smaller. So, to
find the threshold, we must find the point at which the
curves from the different lattice sizes intersect. To
estimate the threshold error rate, we use the method

described by Wang et al. [36] to model the behavior of

the logical error rate close to the crossing. This tells us
that for a large enough lattice size L, the decoding failure
probability is given by
Pfail p pth L1=v0 :

C4

The threshold data are fitted to a quadratic function,

to account for small system-size effects, and the threshold
crossing value drawn from the resultant fit parameters:
Pfail a bp pth L1=0 cp pth 2 L2=0 :

C5

APPENDIX D: COMPUTER
OPERATIONAL SPEED
The entanglement generation described in the previous
section is only one aspect that determines the overall speed
of the device. Figure 8 shows a summary of all factors
contributing to the final clock cycle of the quantum
computer. The contributors are divided into three categories. Entanglement generation has already been discussed.
We now consider the factors affecting GHZ state distillation time and stabilizer measurement.
1. Handling probabilistic stabilizer evaluation
The generation and distillation of entanglement using
optical links between cells is a probabilistic process, and
steps must sometimes be attempted repeatedly. If the cells
do not have sufficient internal complexity to queue up, or
buffer, the results of the different stages of the purification, then necessarily the time taken to complete a
stabilizer measurement will also be probabilistic. In that
case, there is a potential difficulty in performing a
complete set of stabilizers (or a complete subset,
cf. Fig. 7) over the entire computershould we wait
until the very last stabilizer has been successfully
performed, before moving to the next set? In fact, to
do so would require a time cost that scales with the
computer size. Fortunately, this is not necessary; instead,
we can simply wait a fixed time and then abandon any
stabilizers that have not yet been measured. Figure 9
shows the results from recalculating thresholds for the
case where each stabilizer has a 1% chance of not being
evaluated; this shows a minimal difference to the threshold. Limiting the number of stabilizers that must be
completed to a fixed fraction of the whole also keeps the
mean evaluation time constant in the lattice size. We note
that our GHZ-based approach to stabilizer measurement
is particularly friendly to this process of abandoning
the slowest 1% of measurements in each round,
because when we abandon an attempt to create a
GHZ, the data qubits in those cells have not been
involved in any gate operations. Operations on data

041041-10

FREELY SCALABLE QUANTUM TECHNOLOGIES USING

FIG. 8.

The seven major factors contributing to the overall potential running speed of the device.

qubits take place only after successful completion of the

high-fidelity GHZ state.
This approach is still less than ideally efficient since on
average cells will be inactive for a significant portion of
their time, having finished well before the cutoff time.
Since some potential technologies for the cell, including,
for example, NV centers, do not have the internal complexity to act as a sophisticated device such as Fig. 5(c),
it is interesting to ask whether there is another route with
higher efficiency. A possibility is an asynchronous stabilizer measurement, where measurements are made not in
discrete rounds but as soon as the necessary entanglement
has been generated.
2. Cell design to support parallelization
The design of the cell can lead to a number of ways to
parallelize the protocol. The full architecture shown in
Fig. 5(c) is designed to exploit all of these possibilities.

FIG. 9.

PHYS. REV. X 4, 041041 (2014)

Figure 10 shows the relatively modest cells needed for

basic functionality, while Fig. 11 shows cells that boast a
fully parallel architecture.At the lowest level, entanglement generation can be parallelized. If M simultaneous
attempts at entanglement are made, the effective rate of
production is, of course, increased by the same factor.
Further to this, a cell may be able to support entanglement
queuing such that Bell states are continuously created
and stored for later use. In such an architecture, entanglement can be treated as being deterministically generated
at the mean rate (or slightly less, to maintain buffers). On
the other hand, if entanglement must be generated on the
fly, as required, then this must be treated as a stochastic
process. At a higher level, GHZ states can be stored as
and when they are created, leaving other qubits free to
generate more. This removes the inefficiency discussed in
the previous section, where most cells must wait for the
slowest stabilizer measurements to complete. Instead,

Thresholds with 1% of stabilizer outcomes missing: (a) Basic, (b) medium, (c) refined.

041041-11

NICKERSON, FITZSIMONS, AND BENJAMIN

PHYS. REV. X 4, 041041 (2014)

FIG. 10. Four cells of the design shown in Fig. 5(b) with the connections achieved by optical switching. Note that the ion trap elements
in this figure could equivalently be any other few-qubit, optically active system such as a NV center in diamond. The system shown here
is equivalent to that shown in Fig. 11, which has more complex cells and dedicated (nonswitched) cell-cell links; those features increase
the speed, but the error thresholds are the same.

FIG. 11. Four cells of the design shown in Fig. 5(c) with the connections relevant to building their mutual GHZ states highlighted. The
inset shows the abstract concept of of the networked computer from Fig. 1; this ion trap design and the design in Fig. 10 are specific
realizations.

041041-12

FREELY SCALABLE QUANTUM TECHNOLOGIES USING

PHYS. REV. X 4, 041041 (2014)

FIG. 12. Threshold dependence on local error rates for the medium protocol. The threshold for fault tolerance is dependent on all the
physical error rates in the system. In Fig. 4, we show the threshold in the network error rate pn when the local measurement and gate
error rates pm and pg are fixed to 0.05% and 0.1%, respectively. Here, we demonstrate the robustness of the protocol to changes in these
intracell error rates. In (a), we show the thresholds dependence on pg , while (b) reveals the dependence on pm . Each plotted point
represents an entire threshold calculation as shown in the inset of (a). The red data points correspond to the values discussed in the main
text. We stress that as the errors deviate further from this point, it is likely that our protocol becomes suboptimal; thus, the threshold line
should be regarded as a lower bound on the possible threshold.

stabilizers can be measured at the mean rate of GHZ

distillation, providing a further level of speed-up. Finally,
the greater connectivity of the full architecture allows all
the GHZ states required for a stabilizer round (plaquettes
or stars) to be simultaneously distilled. This gives an
additional speed-up factor of 2 over a minimal cell where
this process must be broken into two rounds, as shown
in Fig. 7.

counts is to produce an adjusted level of network noise pn .

From the main text, we have the following expressions for
the case where the dark counts are neglected:
imperfect 1 pn jihj
p
n jihj j00ij00i j11ih11j
3

F1

APPENDIX E: Probing the Threshold

In the main text, we fix the measurement and gate error
rates to pm 0.05% and pg 0.1% [1,2] and then vary
only the network noise to identify a threshold, reporting the
results in Fig. 4. The threshold is, of course, a function
of all the input error rates. We now probe the behavior of
the threshold when these local error parameters are
varied. The threshold is evaluated for several values of
pg using the medium protocol and the method described in
Appendix C 5. The results are shown in Fig. 12. The
threshold exhibits a roughly linear dependence on both pg
and pm , but a stronger dependence on the gate errors is
seen. It is important to note, however, that all threshold
calculations here are made using the same protocol, which
we design to optimize the threshold for the parameters
considered in the main text. For any significantly different
combination of error values, a new protocol designed for
this regime would be likely to outperform the results we
find here.
APPENDIX F: ANALYSIS OF DARK COUNTS
We begin by considering the simple case of r 1=2,
where it is straightforward to show that the effect of dark

and
raw 1 rimperfect rj11ih11j:

F2

Let us take the simple case that r 12, and introduce d

pdc =1 ploss as above, where pdc is the probability that
the system will experience a dark count in a given one of
the two detectors during a given attempt at entanglement.
We can write a new expression 0raw , which has the same
form above, but where pn 0 and r0 replace the unprimed
parameters and have absorbed the dark-count parameter
d. Roughly speaking, pn 0 pn 3d.
Consider the limit of high photon loss, where almost all
dark-count events occur on occasions when all emitted
photons have been lost. When we see a dark count on
such an occasion, we wrongly conclude that we have
heralded the creation of raw . In fact, the state of the two
optically active qubits is simply the completely mixed
state, because they were prepared in an equal superposition
and then (effectively) measured by the environment. Thus,
we can write

041041-13

NICKERSON, FITZSIMONS, AND BENJAMIN

PHYS. REV. X 4, 041041 (2014)



I
raw 2d =1 2d
4


1
1
dI

=1 2d
j11ih11j
2 imperfect 2
2







1 pn d
pn d
pn d 1
1 2d1
jihj
jihj j00ij00i
j11ih11j
2
2
2 2
2
6
6
0
p
1 r0 0imperfect r0 j11ih11j; with 0imperfect 1 p0n jihj n jihj j00ij00i j11ih11j;
3

0raw

where the last line introduces

r0

1
2 4d

and p0n

pn 3d
:
1 4d

F4

We now present a more general analysis of dark

counts for arbitrary r, again demonstrating that they serve
to additively increase the effective network error pn. To
accomplish this, we first note that raw (defined, as in the
main text, as the state heralded by a single click, with dark
counts assumed impossible) is always diagonal in the basis
fji; ji; j00i; j11ig, where ji is the antisymmetric state;
see Eq. (F2). If a nonzero dark-count probability pdc is
taken into account, then the state of the system (given it
has been postselected due to a single detector clicking as
required) will be altered due to three additional ways in
which that single click can be produced. The first is that
the state j00i may survive postselection due to a single
dark count occurring, and this occurs with an absolute
probability p00 2pdc 1 pdc 1 p1 2. The second
way in which a dark count can lead to an effect on
the postselected state is that both a single-photon loss
and a single dark count can occur, resulting in the
state 12 jihj jihj with absolute probability

p 4pdc 1 pdc ploss p1 1 p1 . Finally, a combination of both the loss of two photons and a single dark
count can lead to the erroneous inclusion of the j11i state
after postselection, which occurs with absolute probability
p11 2pdc 1 pdc p2loss p21 . As we wish to examine the
regime where pdc is comparable to or less than the
probability of photon loss not occurring, it is convenient
to introduce the constant d pdc =1 ploss . In the regime
of high loss, where ploss 1, the probabilities become

p00 2d1 ploss 1 pdc 1 r2 ;

F5

p 4d1 ploss 1 pdc r1 r;

F6

p11 2d1 ploss 1 pdc r2 :

F7

Note that in the absence of dark counts the state after

postselection will be raw , which occurs with probability
praw 21 pdc 2 1 ploss p1 1 p1 ploss p1 , which
in the high-loss regime becomes praw 21 pdc 2
1 ploss r. Thus, the state of the system after dark counts
are included is given by

dc

F3

praw raw p00 j00ih00j p11 j11ih11j

2 jihj
praw p00 p11 p

p
2

jihj

1 pdc rraw d1 r2 j00ih00j dr2 j11ih11j dr1 rjihj dr1 rjihj

:
d 1 pdc r

F8
F9

As the dark-count rate in many of the current generation of experiments is already very low, we can consider this expression
in the case of small pdc , in which case
dc

rraw d1 r2 j00ih00j dr2 j11ih11j dr1 rjihj dr1 rjihj

dr

1 rd1 r r p3n 
rd 1r 1 r p3n 

j00ih00j
j11ih11j
dr
dr
r1 rd p3n
r1 rd 1 pn
jihj
jihj:

dr
dr

041041-14

F10
F11

F12

FREELY SCALABLE QUANTUM TECHNOLOGIES USING

PHYS. REV. X 4, 041041 (2014)

When phase drift between the creation of this step and the application of the EPL pair distillation step is taken into account,
the state of the system is given by 0dc 1 pdrift dc pdrift Zdc Z. Hence, we have
0dc

1 rd1 r r p3n 
rd 1r 1 r p3n 
j00ih00j
j11ih11j
dr
dr

F13

r1 rd 1 pn pdrift 1 43 pn 
jihj
dr

F14

r1 rd p3n pdrift 1 43 pn 
jihj:
dr

F15

Note that the application of the EPL protocol conditioned on a (1,1) outcome is not sufficient to ensure that the output
state is in subspace spanned by ji and ji, since this outcome also occurs when one pair is in the state j00i and the
other state is j11i. Note, however, that the (1,1) outcome cannot occur when both pairs are in state j00i or both pairs are in
state j11i. Thus, the pair produced by an application of the EPL protocol to two noisy pairs is
EPL fr; d; pn j00ih00j j11ih11j gr; d; pn ; pdrift jihj hr; d; pn ; pdrift jihj;

F16

where
pn
pn
p1
EPL 1 rd1 r r 3 rd 1r 1 r 3 
j00ih00j j11ih11j;
d r2

F17

pn
4
4
2
2
2
2
p1
EPL r 1 r fd 1 pn pdrift 1 3 pn  d 3 pdrift 1 3 pn  g
jihj;
2
2d r

F18

fr; d; pn
gr; d; pn ; pdrift

hr; d; pn ; pdrift

pn
4
4
2
2
p1
EPL r 1 r d 1 pn pdrift 1 3 pn d 3 pdrift 1 3 pn 
jihj;
d r2

F19

and

FIG. 13. Total error probability. Here, we take r 14 and

pdrift 0.01. The five lines represent the total error probability
corresponding, from bottom to top, to preparation error probabilities pn f0; 0.025; 0.05; 0.075; 0.1g. The error probability
is close to the error rate without dark counts while d is below
approximately 0.01 but increases rapidly thereafter, passing
through the distillation threshold even for the case pn 0.

FIG. 14. Influence of dark counts on effective preparation and

drift error probabilities for r 14 and pdrift 0.01. The five upper
lines (green) represent the effective preparation error weight (p0n )
corresponding, from bottom to top, to preparation error probabilities pn f0; 0.025; 0.05; 0.075; 0.1g. The lower lines (blue),
indistinguishable at this scale, correspond to the effective drift
error weight (pdrift ) for the same values of pn . In the regions
d 0.007 and d 0.469, both p0n and pdrift correspond to
valid probabilities, and hence, in these regimes dark counts
are indistinguishable from other noise sources.

041041-15

NICKERSON, FITZSIMONS, AND BENJAMIN

pEPL

PHYS. REV. X 4, 041041 (2014)

 




r1 r
pn
pn
2pn 2
4 d1 r r

d 1r 1 r
r1 r 2d 1
3
3
3
2d r2

is the probability of obtaining the (1,1) result during

the EPL protocol. The total error probability is then
1 hjEPL ji 1 gr;d;pn ;pdrift . As can be seen
from Fig. 13, dark counts begin to contribute significantly
once d exceeds about 0.01.
It is natural to ask whether the errors present in this state
due to dark counts are fundamentally different from those
due to preparation and drift errors. However, by comparing
EPL to the case of no dark counts, it is possible to find
modified preparation and drift error weights (p0n and p0drift )
such that the two states match. As f is independent
of pdrift , the value of p0n can be obtained by solving
fr; d; pn fr; 0; pn . The value for p0drift can then be
obtained by solving gr; d; pn ; pdrift gr; d; p0n ; p0drift .
Although p0n and p0drift are not guaranteed to correspond
to valid probabilities, for many experimentally relevant
parameter ranges they do indeed take on values between
zero and one, and hence, the effect of dark counts within
these parameter ranges are indistinguishable from drift and
preparation errors. As can be seen from Fig. 14, as long as
the value of d is kept far below 1 (in this case around 0.01),
the effective modification of the drift and preparation error
rates due to dark counts is relatively small. However, as
expected, when d approaches unity the effective error rate
rapidly increases.

[1] T. P. Harty, D. T. C. Allcock, C. J. Ballance, L. Guidoni,

H. A. Janacek, N. M. Linke, D. N. Stacey, and D. M. Lucas,
High-Fidelity Preparation, Gates, Memory, and Readout of
a Trapped-Ion Quantum Bit, Phys. Rev. Lett. 113, 220501
(2014).
[2] C. J. Ballance, T. P. Harty, N. M. Linke, and D. M. Lucas,
High-Fidelity Two-Qubit Quantum Logic Gates Using
Trapped Calcium-43 Ions, arXiv:1406.5473.
[3] R. Barends et al., Logic Gates at the Surface Code Threshold: Superconducting Qubits Poised for Fault-Tolerant
Quantum Computing, arXiv:1402.4848.
[4] F. Dolde, V. Bergholm, Y. Wang, I. Jakobi, B. Naydenov,
S. Pezzagna, J. Meijer, F. Jelezko, P. Neumann, T. SchulteHerbrggen, J. Biamonte, and J. Wrachtrup, High-Fidelity
Spin Entanglement Using Optimal Control, Nat. Commun.
5, 3371 (2014).
[5] T. H. Taminiau, J. Cramer, T. van der Sar, V. V. Dobrovitski,
and R. Hanson, Universal Control and Error Correction in
Multi-Qubit Spin Registers in Diamond, Nat. Nanotechnol.
9, 171 (2014).
[6] C. Monroe, R. Raussendorf, A. Ruthven, K. R. Brown, P.
Maunz, L.-M. Duan, and J. Kim, Large Scale Modular
Quantum Computer Architecture with Atomic Memory and
Photonic Interconnects, Phys. Rev. A 89, 022317 (2014).

F20

[7] D. Hucul, I. V. Inlek, G. Vittorini, C. Crocker, S. Debnath,

S. M. Clark, and C. l. Monroe, Modular Entanglement
of Atomic Qubits Using Both Photons and Phonons,
arXiv:1403.3696.
[8] L. DiCarlo, J. M. Chow, J. M. Gambetta, L. S. Bishop, B. R.
Johnson, D. I. Schuster, J. Majer, A. Blais, L. Frunzio, S. M.
Girvin, and R. J. Schoelkopf, Demonstration of Two-Qubit
Algorithms with a Superconducting Quantum Processor,
Nature (London) 460, 240 (2009).
[9] N. Roch, M. E. Schwartz, F. Motzoi, C. Macklin, R. Vijay,
A. W. Eddins, A. N. Korotkov, K. B. Whaley, M. Sarovar,
and I. Siddiqi, Observation of Measurement-Induced Entanglement and Quantum Trajectories of Remote Superconducting Qubits, Phys. Rev. Lett. 112, 170501 (2014).
[10] H. Bernien, B. Hensen, W. Pfaff, G. Koolstra, M. S. Blok, L.
Robledo, T. H. Taminiau, M. Markham, D. J. Twitchen, L.
Childress, and R. Hanson, Heralded Entanglement between
Solid-State Qubits Separated by 3 Meters, Nature (London)
497, 86 (2013).
[11] W. Pfaff et al., Unconditional Quantum Teleportation
between Distant Solid-State Quantum Bits, Science 345,
532 (2014).
[12] D. Yang, A Simple Proof of Monogamy of Entanglement,
Phys. Lett. A 360, 249 (2006).
[13] W. Dr and H.-J. Briegel, Entanglement Purification for
Quantum Computation, Phys. Rev. Lett. 90, 067901 (2003).
[14] E. T. Campbell, Distributed Quantum-Information Processing with Minimal Local Resources, Phys. Rev. A 76, 040302
(2007).
[15] K. Fujii and K. Yamamoto, Entanglement Purification with
Double Selection, Phys. Rev. A 80, 042308 (2009).
[16] K. Fujii, T. Yamamoto, M. Koashi, and N. Imoto, A
Distributed Architecture for Scalable Quantum Computation with Realistically Noisy Devices, arXiv:1202.6588v1.
[17] M. Zwerger, H. J. Briegel, and W. Dr, Hybrid Architecture
for Encoded Measurement-Based Quantum Computation,
Sci. Rep. 4, 5364 (2014).
[18] N. H. Nickerson, Y. Li, and S. C. Benjamin, Topological
Quantum Computing with a Very Noisy Network and Error
Rates Approaching One Percent, Nat. Commun. 4, 1756
(2013).
[19] A. Kitaev, in Proceedings of the Third International
Conference on Quantum Communication and Measurement
(Plenum Press, New York, 1997).
[20] A. Y. Kitaev, Fault-Tolerant Quantum Computation by
Anyons, Ann. Phys. (Amsterdam) 303, 2 (2003).
[21] E. Dennis, A. Kitaev, A. Landahl, and J. Preskill, Topological Quantum Memory, J. Math. Phys. (N.Y.) 43, 4452
(2002).
[22] C. Horsman, A. G. Fowler, S. Devitt, and R. Van Meter,
Surface Code Quantum Computing by Lattice Surgery,
New J. Phys. 14, 123011 (2012).
[23] S. Bravyi and A. Kitaev, Universal Quantum Computation
with Ideal Clifford Gates and Noisy Ancillas, Phys. Rev. A
71, 022316 (2005).

041041-16

FREELY SCALABLE QUANTUM TECHNOLOGIES USING

[24] A. G. Fowler, A. M. Stephens, and P. Groszkowski, HighThreshold Universal Quantum Computation on the Surface
Code, Phys. Rev. A 80, 052312 (2009).
[25] D. S. Wang, A. G. Fowler, and L. C. L. Hollenberg, Quantum Computing with Nearest Neighbor Interactions and
Error Rates over 1%, Phys. Rev. A 83, 020302(R) (2011).
[26] E. T. Campbell and S. C. Benjamin, Measurement-Based
Entanglement under Conditions of Extreme Photon Loss,
Phys. Rev. Lett. 101, 130502 (2008).
[27] S. D. Barrett and P. Kok, Efficient High-Fidelity Quantum
Computation Using Matter Qubits and Linear Optics,
Phys. Rev. A 71, 060310 (2005).
[28] A. Reiserer, N. Kalb, G. Rempe, and S. Ritter, A Quantum
Gate between a Flying Optical Photon and a Single
Trapped Atom, Nature (London) 508, 237 (2014).
[29] M. M. Steiner, H. M. Meyer, J. Reichel, and M. Khl,
Photon Emission and Absorption of a Single Ion coupled to
an Optical Fiber-Cavity, arXiv:1407.6036.

PHYS. REV. X 4, 041041 (2014)

[30] S. C. Benjamin, D. E. Browne, J. Fitzsimons, and J. J. L.

Morton, Brokered Graph-State Quantum Computation,
New J. Phys. 8, 141 (2006).
[31] S. Bravyi and A. Vargo, Simulation of Rare Events in
Quantum Error Correction, Phys. Rev. A 88, 062308 (2013).
[32] F. H. E. Watson and S. D. Barrett, Logical Error Rate
Scaling of the Toric Code, arXiv:1312.5213v2.
[33] A. Jamiolkowski, Linear Transformations which Preserve
Trace and Positive Semidefiniteness of Operators, Rep.
Math. Phys. 3, 275 (1972).
[34] J. Edmonds, Paths, Trees, and Flowers, Can. J. Math. 17,
449 (1965).
[35] V. Kolmogorov, Blossom V: A New Implementation of a
Minimum Cost Perfect Matching Algorithm, Math. Program. Comput. 1, 43 (2009).
[36] D. S. Wang, A. G. Fowler, A. M. Stephens, and L. C. L.
Hollenberg, Threshold Error Rates for the Toric and Planar
Codes, Quantum Inf. Comput. 10, 456 (2010).

041041-17